KR20080008705A - A street guidance information system for blind peoples that using image recognition technology - Google Patents

Abstract

A street guidance information system for the blind using image recognition technology is provided to process image data on a real time basis and to notify a blind person of the direction or circumstances of a road. A street guidance information system for the blind using image recognition technology includes a central server, a bluetooth device(110) and a central processing device(130). The central server inputs and searches the load information of a starting point and an arriving point, and downloads when the information of the concerned load exists. The bluetooth device is carried by a blind person, includes a CCD camera and an ear stet and contains bluetooth technology. The central processing device is carried by the blind person, and is connected to the bluetooth device via a network or a wire to perceive and process the image.

Description

A Street Guidance Information System for Blind Peoples That Using Image Recognition Technology}

1 is a view showing a visually impaired road guide system using a visual recognition technology according to the present invention.

FIG. 2 is a block diagram illustrating that a visually impaired person downloads image feature values from a central server to a starting point from a central server and stores the image feature values in a visually impaired road guide system using image recognition technology.

3 is a block diagram illustrating a visually impaired road guidance system using image recognition technology according to the present invention.

4 is a block diagram illustrating communication between the Bluetooth device and the CPU of FIG. 3.

5 is a configuration block diagram for analyzing a road image according to the present invention.

6 is a flowchart for performing download by searching road information of a server according to the present invention.

7 is a flowchart for analyzing an image according to the present invention and notifying the result by voice or storing.

Choi Nae-won, Oh Hae-seok, "Object Extraction and Tracking in Real-time Color Image", The KIPS Transactions: PartB, 1598-284X, Vol. 10B, pp.81-86, 2003

Kim, Dong-Wook, "A Study on the Development of Ultrasonic Electronic Walking Inducer to Induce Gait of the Visually Impaired," A Study on the Development of a Ultrasonic Electronic Travel Aid for the Blind, Korea Institute of Science and Technology Information, Vol. 3, No. 4, pp .227-232, 2002

Lim, Jae-Bong, "Inductive Signal Device for the Visually Impaired", Proceedings of the Korean Institute of Communication Sciences Conference, Vol. 97, pp.98-104, 1997

Lee, Kwang-Bok, "System for Providing Location Information for the Visually Impaired Using Radio Frequency Recognition Technology", Patent Application No. 10-2005-0030998

Moin Valley, "Location information service system and service method for the visually impaired", Patent Application No. 10-2000-0056668

Pohang Institute of Industrial Science and Technology, "Electronic sensing wand for visually impaired persons", patent application number 10-2000-0051226

NANUX Co., Ltd., "Eyeglasses with Ultrasonic Sensor for the Visually Impaired", Patent Application No. 10-2000-0062341

UBI Technologies, Inc., "Vibration Voice Guidance System for the Visually Impaired," Patent Application No. 10-2004-0096221

Hyun Chae Jung, Baek Chul Kim, "Directional Voice Output System for the Visually Impaired", Patent Application No. 20-2000-0008558

Cha Bo-young, Chae Hyun-chul, "Frequency Controlled Traffic Signal for Visually Impaired Induction Signal", Patent Application No. 20-2002-0004797

Jeon, Byeong-Soo, Jeon-Oun, "Object Detection System for the Visually Impaired", Patent Application No. 10-1999-0003840

SK Teletech Co., Ltd., "Guidelines for the Visually Impaired Using Ultrasonic Sensors and Mobile Communication Devices Implementing the Same", Patent Application No. 10-2003-0000732

Hangil Handycare Co., Ltd., "Induction Block for the Visually Impaired", Patent Application No. 20-2000-0026476

Choi Heung-tae, "Manufacturing method of the pedestrian guide for visually impaired", Patent Application No. 10-2004-0035654

Pantech & Curitel Co., Ltd., "Pedestrian guide device and method for the visually impaired, mobile phone and personal portable information terminal having a pedestrian guide function for the visually impaired", Patent Application No. 1020040057782

Bon-Kyung Koo, "Wand for the visually impaired with a signal generator", patent application number 1020010027842

Korea Advanced Institute of Science and Technology, "Pedestrian guidance device and method for converting visual information for indoor / outdoor walking of visually impaired person into auditory and tactile information", Patent application No. 1020040100557

LG Electronics Co., "Road guide device and method for visually impaired users of mobile communication terminal", patent application number 1020050019948

Kim Won-taek, "Safety Traffic Light", Patent Application No. 2020010028807

Kim Won-taek, Kim Ji-ho, "Blinking Safety Signal Light", Patent Application No. 2020010028962

Park, Sang-Joon, Yoo, Jung-Yeol, "A device for transmitting position and visual information for the visually impaired by applying barcode pattern and reader", Patent application number 1020040065716

Yuho Ho, Kwon Dae-gyu, Yoon Myung-jong, "Pedestrian guide device for the visually impaired using tactile stimulator and GPS," patent application No. 1020030089236

Chung, Dong-geun, Noh Hyun, "Pedestrian guide device for the visually impaired", patent application number 1020040076563

Pohang Institute of Industrial Science and Technology, "Pedestrian guide device for the visually impaired," Patent Application No. 1019990045992

Minseok Kim, "Guide for the visually impaired", Patent Application No. 1020020050981

Lee In-soo, Yang Il-yong, Park Kwang-cheol, Ko Man-kyu, "Pedestrian guide device for the visually impaired", patent application number 2020050005059

Kim Sun-kyu, "Pedestrian guide device for the visually impaired", Patent application number 2020060002165

Park Myung-sook, "Obstacle detection alarm device for the visually impaired", patent application number 1019990021050

Sungchang Enterprise Co., Ltd., "Guideline service system and method for the visually impaired", Patent Application No. 1020030003751

LG Electronics Co., Ltd., "Device for the visually impaired using mobile terminal", patent application number 1020040079916

SK Telecom Co., Ltd., Kookmin University, "Interactive Guidance Signal System for the Visually Impaired and Its Control Method", Patent Application No. 1019960000845

Korea National University of Technology, Kim Young Joong, Kim Jae Hee, Kim Jung Min, Kyung Jin Jin, Kyu Dong Lee, Cice Syria, "Guide for the visually impaired", Patent Application No. 1020040111969

Seung-Ho Cho, "Location Information Voice Guidance System and Method for the Visually Impaired Using Two-Way Wireless Communication", Patent Application No. 1020030048400

Minchul Kim, "Directional Sound Guidance System", Patent Application No. 1020040017517

Hantec Co., Ltd., Myung-Hee Choi, "Wi-Directional Video Guidance Device for the Visually and Hearing Impaired Persons," Patent Application No. 2020050016694

Eugene Tech Co., Ltd., "Location guide device for the visually impaired", patent application number 2020050027791

Cheol-Ho Yang, "Visual Aids for the Visually Impaired", Patent Application No. 1020010049787

Lee, Sung-Hwan, "Wearable walking guide device and method for the visually impaired", Patent Application No. 1020000027839

Pantech Co., Ltd., "Portable Terminal for Outputting Audio Data Regarding Image Input via Mounted Camera", Patent Application No. 1020040109862

Gigasys Net, "Method and apparatus for converting visual information into auditory information", Patent Application No. 1020020021389

LG Electronics, Inc., "Object Detection Device for Portable Terminal", Patent Application No. 1020010072753

LG Electronics, Inc., "An Obstacle Detection Apparatus and Method for Mobile Communication Terminals", Patent Application No. 1020040069682

Kim, Kyung Hwan, "Walking guide device for the visually impaired", patent application number 2020040025331

LG Electronics, "Mobile Communication Terminal for the Visually Impaired", Patent Application No. 2020000036388

Uhm Hyun-deok, "Road Guidance System", Patent Application No. 2020050021028

Hangil Handycare Co., Ltd., Go Hyun-joo, "Intelligent Traffic Information Voice Guidance System", Patent Application No. 1020020009369

LG Electronics Co., "Geographic Information Providing Device and Method Using Bluetooth Technology", Patent Application No. 1020020005429

LG Electronics Co., "Route guidance method of mobile communication terminal using camera and satellite positioning system", patent application number 1020030022030

Pantech & Curitel Co., Ltd., "Walking Aid Device and Method", Patent Application No. 1020040110114

Pantech Co., Ltd., "Blind Guidance Service Method Using Mobile Communication Terminal", Patent Application No. 1020030089219

Kim Seung-ho, Kim Jae-jun, "Blind Guidance Safety Guidance Device", Patent Application No. 1019990040225

Song Jong-sup, "Reader for the Visually Impaired", Patent Application No. 1020050038189

Jang, Euk-deok, "Blind Image Information Recognition System and Its Control Method", Patent Application No. 1019990060482

SK Telecom Co., Ltd., "Guide for visually impaired", Patent Application No. 1020040023476

UBI Technologies, Inc., "Vibration Voice Information System for the Visually Impaired," Patent Application No. 1020040096221

Gigasys Net, "Method and apparatus for converting visual information into auditory information", Patent Application No. 1020020021389

The present invention receives the image from the starting point to the arrival point from the visually impaired with the help of the first assistant with a CCD camera, and analyzes the image data, such as edge, slope, distance, road direction, location, traffic lights, crossings and milestones. Extract feature values and save them to the final central server. After that, when the visually impaired person wants to go from the same starting point to the arrival point, the image feature value is downloaded from the central server, and then the image is inputted from the starting point to the arrival point in real time by the CCD camera, and the feature value obtained by analyzing the image data is downloaded. The present invention relates to a system and a control method for allowing the visually impaired to arrive safely and conveniently to the arrival point by informing the visually impaired the position and the risk factors to go to the voice.

As of 2005, the number of blind people in Korea has reached 190,000. Even though they should be free to participate in all activities such as education, religion, leisure, shopping, etc. like the non-disabled, there is no place for free activities. In addition, the facilities are inconvenient and unsafe.

The conventional technology is to install a braille block on the sidewalk block to visually impaired people to recognize whether the end of the road, pedestrian crossing, stairs, etc., and more recently has been developed more intelligent using magnetic, ultrasonic or radio signals A walking guidance device has been developed. The technical contents of these are as follows.

The road guidance of the visually impaired by the sidewalk block is only from the braille block, and conventionally, a reader such as a radio wave transmitter or an RFID tag is inserted into the sidewalk block to notify the visually impaired person when trying to leave the block, or a crosswalk or stairs. I'm telling you.

The road guide of the visually impaired by the electronic sensing cane detects an electromagnetic induction block or a short-range obstacle and installs an electronic sensing processing device in the cane to guide the surrounding situation in detail by the electronic sensing processing device installed in the cane. I'm trying to.

Road guidance of the visually impaired by the glasses equipped with the ultrasonic sensor is to detect and guide the obstacle in front of the ultrasonic sensor.

The prior art has a point that must be installed in the braille block of all the roads by using a braille block, a high frequency, an ultrasonic wave, an electromagnetic induction block or the like, or a receiver for indicating the location of the visually impaired. In addition, road guidance using GPS has a problem that it is difficult to apply in practice because the detection error range of the GPS, such as obstacles exceed the error range for walking.

Therefore, an object of the present invention is to process the video data received in real time in real time to recognize the direction of the road or the road situation by voice to the visually impaired, the visually impaired can arrive safely from the starting point to the arrival point To provide a way to do so.

According to an aspect of the present invention in order to achieve the above object, the central server retrieves the image feature value corresponding to the starting point and the arrival point, and downloads and stores in the memory. Images of the road are stored in the input handle through the CCD camera, and these images are stored in the frame transfer buffer unit through the capture unit. The first network detector detects a signal transmission environment of the network and transmits the signal according to the frame transmission buffer unit, and the frame transmission buffer unit outputs the image data stored by the capture unit. The second network detector detects the signal reception environment of the network and transmits the signal according to the frame reception buffer unit, and the frame reception buffer unit transfers the image transmitted to the network to the video output handle to send one image to the video device for each frame. The other is processed and transferred to an image processing unit that recognizes the image. The video output handle outputs an image received from the frame receiving buffer unit to the video device, and the image processor analyzes the image and transmits the result to the result value transmitting buffer unit. The result value transmission buffer unit receiving the image analysis result is notified of the signal transmission environment detection result of the second network detection unit and outputs the result to the network. The first network sensor detects the reception transmission environment of the network and transmits the signal according to the result value receiving buffer unit, and the result value receiving buffer unit stores the image analysis result transmitted to the network and transmits the result to the voice processing unit. The voice processing unit delivers voice data according to the image analysis result to the audio output handle, and the audio output handle provides a visually impaired road guidance system using image recognition technology that notifies the state of the road by audio.

According to another embodiment of the present invention, there is provided a method of downloading and storing image feature value data of a starting point and an arrival point from a central server, temporarily storing an input image signal for each frame, and transmitting environment of a network. Detecting a signal, transmitting a video signal in a set form according to the detected transmission environment, detecting a reception environment of a network, storing a video signal transmitted according to the detected reception environment, and then monitoring each frame through a monitor. Outputting, analyzing the stored image signal for each frame, transmitting the analysis result of the image data in a form set according to the detected transmission environment, storing the image analysis result transmitted according to the detected reception environment, Importing voice data from memory according to the saved image analysis result, imported sound Including the step of outputting data to the audio device to provide a visually impaired road guidance system using an image recognition technology is made.

According to another aspect of the present invention, there is provided a computer readable recording medium tangibly embodying a program consisting of instructions executable by a computer in order to carry out the characteristic method described above.

According to another aspect of the invention, the download unit for downloading the image feature value data corresponding to the starting point and the arrival point from the central server to store in the memory; A capture unit which stores an image signal input through the CCD camera device; A frame transmission buffer unit for dividing the frame into a predetermined size for each frame of the stored image and transmitting the same; A frame receiving buffer unit which receives and stores an image signal applied through a network, divides the image signal into a predetermined size, and outputs the divided image signal; A video device for outputting the received video signal to the screen for each frame; An image processor configured to analyze the received image for each frame, extract feature values, and store a result of comparing the stored data with data stored in a memory; A result value transmission buffer for dividing the result of the image analysis into a predetermined size and transmitting the result; A result value reception buffer unit for receiving and storing an image analysis result signal applied through a network and including at least one reception buffer; A voice processor configured to select voice data corresponding to the stored result value from a memory and store the voice data; An audio device for converting voice data into sound and outputting the sound; The video input handle is alternately received and stored in a plurality of frame buffers and output by a plurality of frame transfer buffer units. The output of the frame receiving buffer unit is alternately received and stored, and the output is alternately applied to the screen. A plurality of image processing units including a plurality of output handles, a plurality of image processing units for extracting feature values by an image processing technique, and a result value obtained by image processing in a plurality of result transfer buffers and then output by a result transfer buffer In addition, there is provided an apparatus for visually impaired road guidance system using an image recognition technology, characterized in that an audio processing unit receives and stores alternately the output of the result value receiving buffer and alternately outputs the output to the earset.

Hereinafter, a road guide system for the visually impaired using the image recognition technology according to the present invention will be described with reference to the accompanying drawings. However, the present invention is not limited or limited by the following examples.

1 is a view showing a blind person equipped with a control device for a visually impaired road guidance system using a video recognition technology according to the present invention, Figure 2 is a block diagram of a configuration for retrieving and downloading data about the starting and destination points from a central server 3 and 4 are block diagrams illustrating a configuration of outputting audio data through image recognition from input image data, FIG. 5 is a block diagram illustrating an image recognition configuration for recognizing an input image, and FIG. FIG. 7 is a flowchart for downloading data stored in the central server, and FIG. 7 is a flowchart of the Bluetooth device from transmitting the input image through the network. FIG. 8 is a flowchart of a central processing unit until receiving image data from a Bluetooth device through a network and transmitting a result through image recognition through a network. FIG. 9 is a flowchart of an image recognition result transmitted from a central processing unit through a network. This is a flowchart for converting and outputting voice data.

As illustrated in FIG. 1, the visually impaired person 140 has a headset 120 of the Bluetooth device 110. , The CCD camera 100 is connected to the Bluetooth device 110, the cane carried by the blind 140 has a central processing unit (130). Here, the blind person 140 walks with the cane 131 having the central processing unit 130, and at this time, an image is input to the CCD camera 100 mounted on the Bluetooth device 110. The input image is transmitted to the central processing unit 130 by the Bluetooth device 110, the Bluetooth device 110 receives the result of the processing and outputs the audio data corresponding to the result to the headset 120.

In the Bluetooth device 110, as illustrated in FIG. 4, the CCD camera device 100, the video input handle 310, the capture unit 320, the frame transfer buffer unit 330, and the first network sensor unit ( 200, the result value receiving buffer unit 430, the voice processing unit 420, the audio output handle 410, and the audio device 120. The central processing unit 130 includes a download unit 700 and a second unit. The network 500 includes a network detector 900, a frame receiving buffer 600, a video output handle 620, a video device 630, an image processor 640, and a result value transmitting buffer 650. By the communication between the Bluetooth device 110 and the central processing unit 130.

Here, the video input handle 310 is composed of an input information unit 311 and an input buffer 312, the capture unit 320 is the first to N-th frame buffer (321-1, 321-2, ...) The frame transmission buffer unit 330 includes first to Nth frame transmission buffers 331-1, 331-2,... The frame reception buffer unit 610-1 includes first to Nth frame reception buffers 611-1, 611-2,..., And the video output handle 620 is a video output information unit 621-1. And a video output buffer 621-2, the image processor 640 comprising a preprocessing 641, a feature value extraction 642, and a stored feature value 643. 1, 640-2, ...). The result value transfer buffer unit 650 includes first to Nth image buffer units 651-1, 651-2,..., 651 -N. The Nth result value reception buffers 431-1, 431-2,..., And the voice processing unit 420 are configured by the comparison voice processing 421-1 and the memory 421-2. The audio output handle 410 may include an audio output information unit 411 and an audio output buffer 412, and the network 500 may include a telephone exchange network, an ISDN (broadband information communication network), or the like.

Referring to the configuration of the visually impaired road guide system using the image recognition technology according to the present invention configured as described above in more detail as follows.

As shown in FIG. 2, the blind person 140 accesses the central server 145 through the Internet, searches for road information 142 corresponding to the starting point and the arrival point 140, and is searched by the search engine 141. The resulting image feature values are downloaded to the memory 143 of the central processing unit 130.

3, the video input information unit 311 in the video input handle 310 of the Bluetooth device 110 continuously detects an image signal input through the CCD camera device 100, and inputs the image. The signal is temporarily stored in the frame buffers 321-1, 321-2,... In the capture unit 320. The first network detector 200 detects a situation of the network 500 to be connected and outputs a detection signal accordingly. The frame transmission buffer unit 330 transmits the number and size of frames of the image frames in the frame buffers 321-1, 321-2, ... by the detection signal output from the first network detector 200. Buffers 331-1, 331-2, ... are prepared and stored.

In addition, the second network detection unit 900 detects the network 500 and outputs a detection signal according thereto, and the frame receiving buffer unit 600 transmits a frame signal transmitted by the Bluetooth device 110 according to the detection signal. Save it. The video output information unit 621-1 of the video output handle 620 sets the size of the video output buffer 621-2 to output through the video device 630, and the frame signal stored in the frame reception buffer unit 600. Is stored in the set video output buffer 621-2 and output through the video device 630. The image processing units 640-1, 640-2,..., 640 -N in the image processing unit 640 preprocess each frame signal stored in the frame receiving buffer unit 600, and extract feature values 642. By extracting the feature value through, and compares it with the stored feature value (643) to check whether the value is valid. The feature values extracted by each image processing (640-1, 640-2, ..., 640-N) are the result value transfer buffer 651-1 in the result value transfer buffer unit 650 by the number and size thereof. 651-2, ..., 651-N). The second network detection unit 900 detects the situation of the network 500 to be connected and outputs a detection signal accordingly. The result value transfer buffer unit 650 is a result stored in the result value transfer buffers 651-1, 651-2,..., 651 -N based on a detection signal output from the second network detector 900. Print the value.

In addition, the first network detector 200 detects the network 500 and outputs a detection signal according thereto, and the result value receiving buffer unit 430 is transmitted by the central processing unit 130 according to the detection signal. Save the result signal. The audio output information unit 411 of the audio output handle 410 sets the size of the audio output buffer 412 to output through the audio device 400. The comparison speech processing 421-1 of the speech processing unit 420 performs the speech and the respective result values stored in the result receiving buffers 431-1, 431-2, ... in the result receiving buffer 430. Compared with the data 421-2, the voice data corresponding to the result value is selected and stored in the audio output buffer 412 in the audio output handle 410. The audio output information unit 411 outputs the voice data stored in the audio output buffer 412 by the audio processor 420 through the audio device 400.

Here, as shown in FIG. 5, the feature values from the image data input to the image processor 640 of FIG. 4 form a basis from which the feature values can be extracted by the preprocessing 641, and then edge extraction 642-1. ), The slope extraction 642-2, the distance extraction 642-3, the height extraction 642-4, the direction extraction 642-5, etc. Extract feature values (642-6) such as traffic lights, crosswalks, and milestones.

The control method of the visually impaired road guidance system using the image recognition technology according to the present invention configured and operated as described above will be described with reference to FIGS. 6 and 7.

First, the blind person 140 connects to the system built in the central server 145 and performs starting point selection (S100) and arrival point selection (S101) in sequence, and whether the central server 145 has data about the same starting point and arrival point. Search (S102). If there is data on the same starting point and arrival point, the image feature values are downloaded (S103, S104) and stored in the memory (S105) of the visually impaired road guidance system using image recognition technology.

Subsequently, the video input handle 311 continuously detects whether there is an image signal input to the CCD camera S200 (S201), and when the video signal is input, temporarily stores the video signal in the input buffer 312 (S202). )do. The video signal stored in the input buffer 312 is separated and stored in each frame by the capture unit 320, and sets the frame transmission buffer according to the size and number of the captured frames (S203). The first network detection unit 200 detects the network 500 situation (S204). The frame signal is stored in the transmission buffer set according to the detection signal of the first network detection unit 200 (S205), and the frame signal of the transmission buffer is detected when the central processing unit and the network are connected (S206). To transmit (S207).

Subsequently, the network 500 is detected by the second network detector 900 (S300), and a frame reception buffer unit for receiving a frame signal transmitted according to the detection signal of the second network detector 900 ( The size and the number of 600 are set (S301), and the received frame signal is stored in the set frame receiving buffers 610-1, 610-2, ... (S302). The received frame signal stored in the frame receiving buffers 611-1, 611-2, ... in the frame receiving buffer unit 600 is stored in the video output buffer 621-2 in the video output handle 620 (S303). The frame signal of the stored video output buffer 621-2 is output to the video device 630 (S304). The received frame signals stored in the frame receiving buffers 611-1, 611-2, ... in the frame receiving buffer unit 600 are subjected to edge extraction (S306) through preprocessing (S305) of the image processor 640. After determining whether the extracted edge information is useful data (S307), the characteristic value of the result of the slope calculation (S308), the distance calculation (S309), the height calculation (S310), and the direction extraction (S311) and the central server 145 Compared with the downloaded data (S312) from the S) how much similarity (S313), and compares whether the current state is the input mode (S314) to determine whether it is the step of storing only the state of the road from the starting point to the arrival point do. In the case of the input mode, the extracted feature value is stored (S315). In the case of the input mode, the extracted feature value is stored in the result value transmission buffer unit 650, and the Bluetooth device and the network 500 are connected. The result stored in the result transmission buffer 650 is transmitted. A result value receiving buffer unit for detecting a network 500 situation by the first network detector 200 (S400) and receiving a result value signal transmitted according to a detection signal of the first network detector 200 ( 430 sets the size and number (S401), and the received result value S402 stored in the set result value receiving buffers 431-1, 431-2, ... is compared with the speech processing in the speech processing unit 420. The audio device 400 stores the voice data S403 corresponding to the result value among the voice data stored in the memory 421-2 by the 421-1 in the audio output buffer 412 in the audio output handle 410. Output to (S404). Thereafter, it is determined whether the extracted feature value arrives at the destination (S317). In case of the destination, the extracted feature value is connected to the central server 145 to store the stored feature value information from the starting point to the arrival point (S318).

In the system and method for visually impaired road guide system and method using the image recognition technology according to the present invention configured and operated as described above, the visually impaired person sets a starting point and an arrival point, downloads a corresponding video feature value from a central server, and stores it in a memory. From the starting point to the arrival point, the road image is input to the CCD camera, the input image data is analyzed, the feature value is extracted, and the voice data corresponding to the result value is selected and informed to the visually impaired person.

The visually impaired road guidance system using the image recognition technology according to the present invention is a system for visually impaired people to safely arrive from the starting point to the arrival point by analyzing the incoming video in real time, the situation of the road by immediately analyzing the incoming video data in real time In addition, there is an effect that the blind person can walk safely by performing obstacles, whether there is construction, whether the direction is presented, traffic light notifications, crossing notices and milestones.

In addition, there is an effect that the visually impaired can immediately deal with the visually impaired by telling the visually impaired or the situation of the road.

Those skilled in the art will appreciate that various modifications and changes can be made in the present invention without departing from the spirit and scope of the invention as set forth in the claims below.

Claims (23)

In the visually impaired road guidance system using image recognition technology, A central server capable of inputting and searching road information of a starting point and an arrival point and downloading the corresponding road information if there is information on the corresponding road; A Bluetooth device which is carried by the visually impaired and has a CCD camera and an earset and which contains Bluetooth technology; A central processing unit which is carried by the visually impaired and which is connected to the Bluetooth device by a network or a wire and can recognize and process an image; The central server; Providing a web site and a database for the visually impaired to store, enter, and retrieve data through the Internet, The Bluetooth device which is a portable device for the visually impaired; It transmits the video signal input by CCD camera to the central processing unit wirelessly or wired through the network on a frame-by-frame basis, and provides the audio data according to the received image recognition result through the voice device. The central processing unit is a portable device for the visually impaired; Receives a video signal in the unit of frame transmitted through the network and outputs one side through a video device, and the other side extracts the frame video signal from a slope extraction, a distance extraction, a height extraction, a direction extraction, a traffic light extraction, a crosswalk extraction, and a milestone. A visually impaired road guidance system using image recognition technology, characterized in that the image recognition processing such as extraction and transmits the result over a network. The method of claim 1, The central server is configured to allow the visually impaired to store road image feature value data of a starting point and an arrival point through the Internet, retrieve the stored data, and download necessary data through the Internet. The method of claim 1, The Bluetooth device, A video input handle for temporarily storing an image signal input by the CCD camera; A first network detector for detecting a signal transmission state of the network and outputting a signal accordingly; A frame transmission buffer unit for setting the size and number of frame buffers for storing the image signals stored in the video input handle according to the detected signal transmission situation, and storing the image signals in the set frame buffer and outputting the image signals; A result value receiving buffer unit for setting a size and number of buffers for storing result values transmitted through the network according to a signal reception situation of the network and storing the result values in the set buffer; A voice processor which replaces the result value output from the result value receiving buffer unit with a voice stored in a memory and stores the result value in a handle for outputting a voice; And a plurality of audio output handles for outputting the voice stored by the voice processor. The method of claim 3, The video input handle includes a video input information unit for detecting information of the input video signal, and a video input buffer for temporarily storing the video signal output from the input information unit. The method of claim 3, The frame transmission buffer unit is composed of a plurality of frame transmission buffers of which size and number are set according to the detected transmission situation. The method of claim 3. The result value receiving buffer unit is configured to set the size and number of the buffer for storing the result value transmitted through the network in accordance with the detected reception situation, and the Bluetooth device configured as a result value receiving buffer unit for storing the set result value . The method of claim 3, And a comparison voice process for replacing the result value output from the result value receiving buffer unit with voice data stored in a memory and outputting the result value to an audio output handle, and a memory in which voice data according to each result value is stored. The method of claim 3. The audio output handle may include an output information unit configured to detect output information for outputting an audio signal, which is replaced with voice data stored in the memory, through a speaker or earphone, and an audio signal having a size set according to the detected output information. Bluetooth device configured as an output buffer which is then output through the speaker or earphone. The method of claim 1, The central processing unit, A second network detector configured to detect a reception state of the network and output a signal according to the received situation; A frame receiving buffer unit configured to set a size and number of a frame receiving buffer for storing frame signals transmitted through a network according to the detected signal reception situation, and to store frame signals in the set frame receiving buffer; A video output handle for storing and outputting a frame signal output from the frame receiving buffer unit; An image processor to image-recognize the frame signal output from the frame reception buffer unit, extract a feature value, and compare the feature value with stored road information and output a result value; And setting a size and number of buffers for storing result values output from the image processor according to the detected signal transmission situation, and including a result value transmission buffer unit for outputting the resultant values in the set buffer. Central processing unit. The method of claim 9, And the frame receiving buffer unit comprises a plurality of buffers whose size and number are set according to the detected reception situation. The method of claim 9, The video output handle may include a video output information unit configured to detect output information for outputting a frame signal output from the frame reception buffer unit through a screen, and a video signal having a size set according to the detected output information. Central processing unit consisting of a video output buffer to output through. The method of claim 9, The image processing unit performs a preprocessing process for image recognition processing of the frame signal output from the frame receiving buffer unit, and then extracts the feature value and compares it with the feature value downloaded and stored by the blind person from the central server. A central processing unit configured to store the result in a result transfer buffer. The method of claim 9, And the result value transfer buffer unit comprises a plurality of buffers whose size and number are set according to the detected transmission situation. In the control method of the visually impaired road guidance system using image recognition technology, Temporarily storing an input video signal; Detecting a transmission status of the network; Setting a size and number of the transmission buffers according to the detected transmission situation; Storing and transmitting the audio signal in a transmission buffer in which the size and number are set; Detecting a reception state of the network; Setting a size and a number of the reception buffers according to the detected reception situation; Storing the received frame signal in a frame receiving buffer in which the size and number are set and outputting the same through a screen; Storing the received frame signal in a frame receiving buffer in which the size and number are set and extracting feature values through image recognition processing; Comparing the extracted feature values with road information stored in a central server; Storing and transmitting the resultant value of the extracted feature value and the road information in a transmission buffer in which the size and number are set; Storing the received result value in a result value receiving buffer in which the size and number are set, and outputting a voice corresponding to the result value through an audio device among voice data stored in a memory; The method of claim 14, The transmitting of the frame signal may include setting the transmission size and the number of times of the frame signal according to the detected transmission situation. The method of claim 14, The method of controlling the visually impaired road guidance system using image recognition technology, wherein the transmitting and receiving situation detection steps are performed synchronously. The image recognition processing step is a control method of the visually impaired road guide system using an image recognition technology performed by a thread (Thread) for each frame. The image recognition processing step is a pre-processing process for each frame, the control method of the visually impaired road guide system using image recognition technology to extract the feature value by extracting the edge, slope, distance, height and direction. 19. A computer readable recording medium tangibly embodying a program consisting of instructions executable by a computer to perform the method of any one of claims 15-18. In the visually impaired road guidance system using the image recognition technology that the user terminal is coupled via a network, The user terminal, A frame receiving buffer unit which receives, stores and divides a video signal applied through the network into a predetermined size and includes at least one frame receiving buffer; A video device for converting and outputting a frame signal into an image; A plurality of output handles that alternately receive and store the output of the frame receiving buffer unit and apply the output to the screen; An image processor extracting a feature value of each frame by performing image recognition on each frame of the frame receiving buffer unit; A download unit for storing feature value data of an image corresponding to a starting point and an arrival point from the central server in a memory and comparing the extracted feature value with the extracted feature value; A result value reception buffer unit which receives and stores the result value extracted by the image processing unit applied through the network and divides the result value into a predetermined size and includes at least one result value reception buffer; And a plurality of output handles which receive and store voice data stored in a memory corresponding to a result value received in the result value reception buffer unit and alternately output the voice data to the speaker or the earphone. A visually impaired road guidance system using image recognition technology. The method of claim 20, Receiving and storing the plurality of frame output handles and the output operation of the plurality of voice data output handles is performed by a separate thread operating independently, the visually impaired road guidance system using the image recognition technology. The method of claim 20, The user terminal may further include a network detector configured to monitor the network situation and to control the size and number of the reception buffers included in the frame reception buffer unit and the result value reception buffer unit based on the monitoring result. Visually impaired road guidance system using image recognition technology. The method of claim 20, The feature value is extracted by preprocessing each frame received in the frame receiving buffer unit and recognition processing of edge, slope, distance, height, direction, traffic light, crosswalk, and milestone, and compared with the downloaded image feature value. The visually impaired road guidance system using image recognition technology, characterized in that for outputting the audio data corresponding to the value to the speaker and earphone by the audio device.

Images